Teach the computer your own hastas. Hold each one in front of the camera a few times, and it learns how your hand looks — then the Hasta Performer can recognise it live.
1Pick a hasta from the list — 28 asamyuta (one hand) and 24 samyuta (both hands).
2Press record. You get a 5-second countdown to form the hasta and hold still.
3Record 3–5 samples of each, then export your model and open it in the Hasta Performer.
Needs a camera and works best in Chrome. Nothing is uploaded — everything stays on your device.